Hertizan fracture theory, previously developed for ideal brittle solids fractured under essentially environment‐free conditions, is here modified to allow for an interaction between the growing cone crack and a reactive environment. Two alternative models, both of which account for the observed detrimental effect of an environment on the Hertzian strength but which predict slight differences in the growth of the cone crack, are offered. The first is based on a surface‐energy‐lowering concept and the second on a subcritical‐crack‐growth concept. The theoretical implications of these two models are examined in the light of Hertzian fracture experiments on abraded glass slabs in different test environments. Observations of the growth of the cone crack in fatigue experiments favor the second model for glass. The advantages of the Hertzian test as a tool for the study of environmental effects on the fracture strength of brittle solids in general are discussed in relation to more conventional fracture tests.